This proposal addresses the need to accurately model, during brachytherapy treatment planning, the often large dosimetric effects of tissue and applicator heterogeneities. Our previously-funded work shows i) that heterogeneity corrections below Cs-137 energies can vary significantly (by as much as 1.5- to 6-fold) with the 3D geometry of the problem and ii) that Monte Carlo photon (MCPT) transport simulation accurately (3-5%) models these effects throughout the brachytherapy energy range. Finding i) precludes using simple one-dimensional calculational algorithms and makes more complex semi-empirical modeling very difficult. Although MCPT is now accepted as a useful and accurate dosimetry tool, it is simply too slow to support patient-specific treatment planning in its current form. The principal project aim is to adapt numerical solutions of the fundamental Boltzmann photon transport equation to accurately (5%) calculation 3D dose distributions in the presence of complex heterogeneous geometries characteristic of typical brachytherapy patients. Two classes of candidate solutions will be investigated for use as archival treatment planning tools (approximately 8 hours computing time). This approach is justified by the success of transport calculations in modeling heterogeneities and the dramatic growth in affordable computer power. Using novel variance reduction and parallel processing techniques, we propose to accelerate forward MCPT simulations 100-fold. Secondly, we will investigate the 3D discrete ordinates method, an efficient (50-fold) deterministic solution not previously investigated in the low-energy photon, shallow- penetration regime. Finally, we propose to complete development of our most promising semi-empirical algorithm, the scatter subtraction model. It will support more rapid (approximately l hour) but more approximate (approximately 10% accuracy) patient dose calculations. Our second aim is to evaluate the universally-held assumption that transport of secondary charged particles can be neglected in brachytherapy. Both measurements and transport calculations will be used to study breakdown of secondary charged particle equilibrium (l) in the presence of steep dose gradients near sources; (2) within thin dosimeters; and (3) near metal-tissue interfaces. Thirdly, we propose to utilize dual-energy CT-scanning to define the composition, geometric architecture and dosimetric consequences of soft tissue, bone and air-cavity heterogeneities in and around common human implant sites.